# Nutrafol Co-Founder Urges Personal Liability for AI Safety Risks By Riza Dagoc Canonical URL: https://www.tokenpost.com/news/regulation/34066 Published: 2026-10-17T14:54:53.000Z Updated: 2026-10-17T14:54:53.000Z Section: Regulation > S. 5576 would require developers to name a safety-plan officer and make developers liable for up to $250,000 per violation. Nutrafol co-founder and AI investor Giorgos Tsetis is urging personal legal consequences for executives who knowingly accept serious public risks from artificial intelligence, as U.S. proposals place formal duties on companies. His argument focuses on individual decisions after risks are known, while a Senate bill would assign safety-plan responsibility to a named officer and penalties to the developer. In an Oct. 17 commentary published at 7 a.m. ET (11:00 UTC), Tsetis said executives should face consequences if they release systems despite known risks, cut safety resources, ignore warnings or conceal problems. “When the people at the top know the risks, what they do next should carry real personal stakes,” he wrote. He left it to lawmakers and other experts to decide whether the threshold should bring civil or criminal consequences. A White House accord dated Sept. 29 calls on participating companies to establish internal controls, empower an internal monitoring team, obtain independent external assessments and provide oversight through an independent board committee. The accord says the measures could later be written into law or regulation, but it does not establish penalties. It follows the administration’s [earlier voluntary AI safety push](). Sen. Mark R. Warner introduced S. 5576, the Artificial Intelligence Risk Management and Security Act of 2026, on Sept. 29. The proposal would create an Artificial Intelligence Safety Board and require developers to file model safety plans that identify the corporate officer responsible for implementation. The bill’s penalty provision would make a developer liable for up to $250,000 per violation of standards adopted under the proposal; it does not specify a separate personal penalty for the named officer. The bill was read twice and referred to the Senate Committee on Commerce, Science, and Transportation. Its proposed developer penalties apply to violations of standards adopted under the bill, rather than to the officer’s designation alone. Joonko Diversity’s founder and former CEO pleaded guilty to securities fraud on Sept. 11 over misrepresentations about the company’s customers and revenue. That case involved investor fraud, not harm caused by an AI system, and does not resolve what liability executives face for AI safety decisions. ## Links in this article - [earlier voluntary AI safety push](https://www.tokenpost.com/news/regulation/25525)